Geth rakudo/nom: e2ec569b7c | (Zoffix Znet)++ | src/core/IO/Handle.pm
Implement IO::Handle.slurp(:bin)

During IO::Grant, we removed :bin from .slurp-rest, because it mutated the handle's encoding. After the Deasyncing of IO, we clarified the spec in that it was allowed to use binary read methods on non-bin handles. With that in mind, re-adding a non-$!encoding-mutating :bin to .slurp aligns the interface with other methods and makes it easier to bin-slurp from Proc pipes too.		 00:46
roast: 1aa8a8ad18 | (Zoffix Znet)++ | S32-io/slurp.t
Spec IO::Handle.slurp(:bin)

Rakudo impl: github.com/rakudo/rakudo/commit/e2ec569b7c
MasterDuke Zoffix: i have a suggestion for synopsebot. if the sender is Geth, maybe it should wait until it sees a 'review: <link>' before reporting the RTs. right now it sometimes interrupts long commit messages 00:59
Zoffix ok 01:02
.in 21h geth synopsebot
yoleaux Zoffix: I'll remind you at 22:02Z
Geth rakudo/nom: be83cd4ec7 | (Zoffix Znet)++ | src/core/Buf.pm
Make Blob.gist 40x faster

Not exactly a method that gotta be über fast, but the original takes 50ms per single call on a 100+ el buf (for comparison 100+ list takes 3ms)		 01:39
Zoffix $bed.append: $Zoffix
on second thought..... 01:53
Geth ¦ rakudo/nom: 66a01aa827 | (Zoffix Znet)++ | src/core/Buf.pm
travis-ci Rakudo build failed. Zoffix Znet 'Revert "Make Blob.gist 40x faster" 03:50
travis-ci.org/rakudo/rakudo/builds/283989608 github.com/rakudo/rakudo/compare/b...a01aa827c1
buggable [travis build above] ☠ All failures are due to: failed make test (1 failure). Across all jobs, only t/04-nativecall/13-cpp-mangling.t test file failed.
[Tux] This is Rakudo version 2017.09-217-g66a01aa82 built on MoarVM version 2017.09.1-553-ga4fef0bd 06:39
csv-ip5xs 1.230 - 1.242
test 9.829 - 9.919
test-t 3.255 - 3.319
csv-parser 0.863 - 0.890
CSV::Parser still br0ken
samcv .tell lizmat i fixed RT132216, where quaternary level was broken in moar blead 07:44
yoleaux samcv: I'll pass your message to lizmat.
synopsebot RT#132216 [new]: rt.perl.org/Ticket/Display.html?id=132216 [UNI] [TESTNEEDED] 'a' coll 'A" not Same but More with disabled tertiary and primary levels
Zoffix squashable6: status 08:50
squashable6 Zoffix, ⚠🍕 Next SQUASHathon in ≈1 hour (2017-10-07 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
Zoffix 1 hour... 08:51
Geth roast: c9af640c46 | pmurias++ | S02-names/is_default.t
Remove fudged tests for a is default(*) on a native variable

is default isn't supposed to work on natives		 09:12
roast: 73b10388fb | (Samantha McVey)++ | S32-str/Collation.t
Add collation tests, testing with $*COLLATION

Should add tests to close RT#132216 along with others. Added tag
  [6.d-REVIEW] in a comment, since it should be reviewed before
a 6.d branch is created.		 09:22
synopsebot RT#132216 [new]: rt.perl.org/Ticket/Display.html?id=132216 [UNI] [TESTNEEDED] 'a' coll 'A" not Same but More with disabled tertiary and primary levels
Geth rakudo/nom: 96669c6b0b | (Samantha McVey)++ | t/spectest.data
Add S32-str/Collation.t to spectest.data
 09:24
Zoffix ZOFFLOP: t/spec/S11-modules/nested.t 10:09
Geth rakudo/nom: 20a99fc3ae | (Zoffix Znet)++ | src/core/Buf.pm
Make Blob.gist 26x faster
 10:10
rakudo/nom: aad8991e2a | (Zoffix Znet)++ | src/core/Map.pm
Fix Map.perl losing cont / Make Map.gist trim to 100els

Hash, List, and Buf trim .gist to 100 els, do the same for Map		 10:14
rakudo/nom: 39461368b5 | (Zoffix Znet)++ | src/core/Hash.pm
Make Hash.perl/.gist use proper object name in circularities
 10:15
rakudo/nom: 69af24c49e | (Zoffix Znet)++ | src/core/Hash.pm
Make Hash.gist 24% faster
Zoffix Will commit tests for the above things that need tests in ~12hr 10:16
Zoffix &
timotimo i wonder if .base can be expressed in a simple nqp op that ends up another 10x faster
Zoffix timotimo: maybe. I just wasn't sure what's the range of typification a Blob's elements could be 10:17
m: Blob[num32].new(2e0) # huh?
camelia concatenate requires a concrete string, but got null
in any protect at gen/moar/stage2/NQPCORE.setting line 1033
in block <unit> at <tmp> line 1
travis-ci Rakudo build passed. Samantha McVey 'Add S32-str/Collation.t to spectest.data' 10:18
travis-ci.org/rakudo/rakudo/builds/284116894 github.com/rakudo/rakudo/compare/6...669c6b0b98
timotimo what, we support num bufs?
Zoffix timotimo: don't know, hence the .base and not an op :) From role parametarization, I'd guess we support anything a native array can hold 10:35
m: use nqp; class Z is repr("VMArray") is array_type(num32) { method new (*@_) { nqp::push(self, $_) for @_ } }.new(2e0, 3e0).say 10:38
camelia (signal SEGV)
Zoffix Weeeeee \o/
timotimo whoops. 10:39
Zoffix m: use nqp; nqp::push(class :: is repr("VMArray") {}, 42) 10:40
camelia (signal SEGV)
Zoffix (no .new)
timotimo ok, that should be easy to catch
Zoffix Filed as rt.perl.org/Ticket/Display.html?id=132228 10:41
Zoffix &
timotimo it'd be so cool if we had a moar built with debug symbols on camelia and on a segfault it'd immediately spit out the backtrace or something 10:45
the push opcode doesn't even check concreteness, even though it probably should 10:47
timotimo AFK
Geth nqp: cb91d63b5f | pmurias++ | 2 files
[js] Throw an exception when we can't find an attribute
 11:03
nqp: 3a01bcdd0f | pmurias++ | t/nqp/058-attrs.t
Test that we throw an exception for missing attributes with nqp::{get,bind}attr
nqp: 759496d50f | pmurias++ | src/vm/js/nqp-runtime/reprs.js
[js] Make MultiDimArray compose be able to take Perl 6 level ints
travis-ci Rakudo build passed. Zoffix Znet 'Make Blob.gist 26x faster' 11:09
travis-ci.org/rakudo/rakudo/builds/284134258 github.com/rakudo/rakudo/compare/9...a99fc3aef0
lizmat Files=1226, Tests=75559, 348 wallclock secs (14.86 usr 5.43 sys + 2443.96 cusr 226.24 csys = 2690.49 CPU) 11:15
yoleaux 06:22Z <teatime> lizmat: the link "the issues at hand" on latest weekly is broken.
07:44Z <samcv> lizmat: i fixed RT132216, where quaternary level was broken in moar blead
09:05Z <teatime> lizmat: looks like because I'm not logged in; this works even if logged out: github.com/search?q=is%3Aopen+is%3...ktoberfest
synopsebot RT#132216 [new]: rt.perl.org/Ticket/Display.html?id=132216 [UNI] [TESTNEEDED] 'a' coll 'A" not Same but More with disabled tertiary and primary levels
lizmat samcv++ 11:17
teatime++ # link fixed 11:18
travis-ci NQP build failed. pmurias '[js] Make MultiDimArray compose be able to take Perl 6 level ints'
travis-ci.org/perl6/nqp/builds/284152755 github.com/perl6/nqp/compare/1e175...9496d50f2f
Geth rakudo/nom: d565ded029 | (Elizabeth Mattijsen)++ | 5 files
Make sure we have IO::Special before IO::Handle

Needed for an optimization later		 11:21
pmurias the dies-ok problem we had with the divide by zero has reappeared on the nqp-js :( 11:33
Geth roast: 7de1138c34 | (Zoffix Znet)++ | S32-hash/perl.t
Map.perl preserves Map's scalar containeration
 11:56
Zoffix Can someone with keys add `#perl6` to geth's web hook for the duration of squashathon? 11:57
And to moarvm
huggable: geth 11:58
huggable Zoffix, Set geth report URL (select application/json and send me everything): geth.niner.name:8888/?chan=#perl6-dev Old URL: geth.perl6.party/?chan=#zofbot
Zoffix So it'd be geth.niner.name:8888/?chan=#perl6-dev,#perl6
Geth roast: 7de1138c34 | (Zoffix Znet)++ | S32-hash/perl.t
Map.perl preserves Map's scalar containeration
rakudo/nom: 381c4c3bb2 | (Zoffix Znet)++ | src/core/Map.pm
Add space between els in Map.gist

We add one in Hash, List, and Buf .gists
  (Noticing though that Buf and List gists separate els with spaces,
but Hash/Map with commas + spaces)		 12:24
roast: 8a3908c3e2 | (Zoffix Znet)++ | S32-hash/map.t
Spec Map.gist trimming to 100 els
 12:25
stmuk . 12:41
Zoffix :
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Make sure we have IO::Special before IO::Handle 12:52
travis-ci.org/rakudo/rakudo/builds/284158820 github.com/rakudo/rakudo/compare/6...65ded02928
Geth rakudo/nom: b377de1c48 | (Zoffix Znet)++ | src/core/Exception.pm
Versatilize X::Numeric::CannotConvert a bit

Make it stringify .defined $!targets as is, while .perl'ifying others. This maintains the old output for X::Numeric::Real, while leaving open the possibility of using more than just a type for
  $!target if we'll need to.		 13:01
dogbert17 .seen CallMeBob 13:31
yoleaux I saw CallMeBob 5 Oct 2017 12:19Z in #perl6: <CallMeBob> -_- 13:32
Zoffix ? 13:35
dogbert17: whatdyuwant? 13:36
dogbert17 just wondering if that's your new alias :) 13:41
Zoffix None 13:42
I setup yoleaux to remind me of a bunch of stuff in /msg and I realized I won't get the reminders 'cause it won't sent them to the right person
dogbert17 aha 13:45
AlexDaniel . 14:32
Geth nqp: 60f79d38e9 | (Zoffix Znet)++ | tools/build/MOAR_REVISION
Bump MoarVM
 15:30
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...5-gd4e230a
rakudo/nom: c4b96e457b | (Zoffix Znet)++ | tools/build/NQP_REVISION
Bump NQP
rakudo/nom: version bump brought these changes: github.com/perl6/nqp/compare/2017....9-g60f79d3
8ff76b5962 | (Elizabeth Mattijsen)++ | 27 files
Zoffix New blog post: "CPAN6 Is Here": 6lang.party/post/CPAN6-Is-Here 15:33
lizmat Zoffix++
jdv79 pause/cpan support was there since at least 2015 if not earlie 15:36
the only thing that happened at that summit was the auto detection of p6 dists
but its not important
cool article though 15:37
travis-ci NQP build failed. Zoffix Znet 'Bump MoarVM' 15:50
travis-ci.org/perl6/nqp/builds/284262049 github.com/perl6/nqp/compare/75949...f79d38e999
Zoffix Same failure as previous bumps: [JVM] t/nqp/058-attrs.t ...................... 1/29 java.lang.ClassCastException: org.perl6.nqp.sixmodel.reprs.P6strInstance cannot be cast to java.lang.String 15:53
samcv yay did mvm get bumped? 16:58
timotimo looks just like 17:00
Zoffix Yes, though there's still a bugglet with t/spec/S17-channel/stress.t It alone takes almost twice as long as entire stresstest now
timotimo Zoffix: the <abbr> for PAUSE on your blog post has a bug, it has a closing ] but no opening [ 17:10
(and why PAUSE has an abbr, but CPAN does not ... ?)
Zoffix Thanks, fixed the typo 17:13
CPAN is more well-known. If people don't know what it stands for they can google it 17:14
timotimo Zoffix, you think it's safe to kick JSON::Fast off of the p6c list?
it's kind of a module that's used by a pretty big portion of the ecosystem
japhb CPAN also doesn't collide with a common English word
timotimo wouldn't want people to experience issues
oh, that's a good point, japhb 17:15
Zoffix timotimo: I think it's safe.
timotimo the more dists move to cpan, the faster the modules page can update, right?
Zoffix Yes. CPAN processor is super fast 17:16
timotimo i have four dists on there, i'll look into making the others cpan-ready as well
every module moved gives us about 5 seconds?
japhb Zoffix: typo: 'nttp' (the link is correct, just the anchor text is wrong)
Zoffix Yeah, something around that
japhb Why is the CPAN processor so much faster? 17:17
timotimo we're not allowed many requests to github per minute
samcv about to remove ornate parenthesis from nqp in a sec
japhb Ah!
Zoffix japhb: thanks. Fixed
timotimo samcv++
japhb np
Zoffix japhb: also we have to make requests in the first place. CPAN processor just rsyncs everything and then reads off the local disk, whereas GitHub processor has to fetch meta file off GitHub (while staying within limits), fetch commit info for freshly updated lists, and also has META validator check the source URL in the META file is a valid URL. 17:19
There's room to improve rate limiter, but that won't make it as fast as CPAN processor: github.com/perl6/modules.perl6.org/issues/79 17:20
japhb wonders if just keeping local mirrors of the GitHub repos would be faster than using their API 17:21
Geth nqp: 02a426e0e6 | (Samantha McVey)++ | src/HLL/Grammar.nqp
Remove support for ornate parenthesis from quoting constructs

Reasons:
  * These are not recognized by Unicode as being brackets
  * They do not conform to Pi/Pf or Ps/Pf general category pairings
  * They don't have a Bidi_Mirroring_Glyph or a Bidi_Matching_Bracket pair
  * They have never been considered brackets for any of the history of
   Perl 6, and they never will be considered brackets
  * There are no modules which use these brackets, and no known users
  * It is not easily versionable, and not feasible to make it versioned
samcv there we go. i think i listed enough reasons
samcv reason 7: samcv doesn't like them :P
timotimo Zoffix: i guess the first step would be to just query that for every 100 requests and output it in the log so we can get an idea of how far we're away from max allowed pace 17:22
gaaah, i want to write code but i have to be careful with my wrists :( 17:23
Zoffix :(
japhb timotimo: Been there. REALLY HATED THAT. 17:26
Geth nqp: 576d78eef8 | (Samantha McVey)++ | src/HLL/Grammar.nqp
Add a removal message if ornate parentheses are used

This message will eventually be removed, but in the very unlikely case anybody is using them, provide a useful message.		 17:34
samcv cool. now nobody will have to come and ask why they code broke. just in the very very unlikely case someone out there *is* using ornate parens 17:35
and i don't think it should slow down parsing since that should only run if we can't find matching parens
Zoffix But doesn't that branch run when non-matching delimiters are used 17:40
m: q:meows:
camelia 5===SORRY!5=== Error while compiling <tmp>
Colons may not be used to delimit quoting constructs
at <tmp>:1
------> 3q:meows:7⏏5<EOL>
expecting any of:
colon pair (restricted)
Zoffix m: q'meows'.say # like this case, for example 17:41
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3q'meows7⏏5'.say # like this case, for example
expecting any of:
infix
infix stopper
statement end
statement mo…
Zoffix was expecting that to work :(
samcv hmm it might
Zoffix m: q|meows|.say # like this case, for example
camelia meows
samcv well worse case we get imperceptibly slower parsing for a month or two 17:42
Zoffix .in 3m time to toss this: github.com/perl6/nqp/commit/576d78eef8
yoleaux Zoffix: I'll remind you at 17:45Z
samcv since i made indexing a single grapheme 9x faster in the last few months :p
Zoffix uhhh
.in 90d time to toss this: github.com/perl6/nqp/commit/576d78eef8
yoleaux Zoffix: I'll remind you on 4 Jan 2018 17:42Z
Zoffix \o/
samcv nice! 17:43
yoleaux Zoffix: time to toss this: github.com/perl6/nqp/commit/576d78eef8 17:45
Zoffix yoleaux: bad robot! 17:46
samcv welp 18:06
samcv goes to toss
Geth roast: 7fd6b6f3ac | (Moritz Lenz)++ | S04-statements/gather.t
Test for RT #127672

take inside a regex with m:g// runs the expected number of times		 18:10
synopsebot RT#127672 [new]: rt.perl.org/Ticket/Display.html?id=127672 ｢take｣ inside global and exhaustive matching is broken (m:g/. {take 1}/)
Geth roast: smls++ created pull request #341:
Add tests for backtracking into alternations		 19:55
roast: 65a762217c | smls++ | 2 files
Add tests for backtracking into alternations

They're the bottom eight tests listed at
  github.com/perl6/nqp/pull/368
  (The other ones listed there, for quantifiers,
seem to be already covered in S05-mass/rx.t)
This closes RT #130117 and #131973.		 19:56
roast: b1ba026b46 | lizmat++ (committed using GitHub Web editor) | 2 files
Merge pull request #341 from smls/master

Add tests for backtracking into alternations
synopsebot RT#130117 [open]: rt.perl.org/Ticket/Display.html?id=130117 [TESTNEEDED] [REGEX] Sequential alternation `||` does not respect `:ratchet`
synopsebot RT#131973 [open]: rt.perl.org/Ticket/Display.html?id=131973 [TESTNEEDED] [REGEX] Backtracking modifiers on individual atoms fail to override a regex-global `:ratchet` modifier.
roast: 946d2ff79d | (Samantha McVey)++ | 2 files
Add ignorecase and ignoremark tests to cover interpolation

Add tests to cover RT132232 and RT132233 which I have just opened to track the interpolation issues with ignorecase. There may be issues with ignoremark or ignorecase+ignoremark as well.		 20:37
synopsebot RT#132232 [new]: rt.perl.org/Ticket/Display.html?id=132232 [UNI] ignorecase returns too many graphemes for expanding foldcase graphemes
RT#132233 [new]: rt.perl.org/Ticket/Display.html?id=132233 [UNI] ignorecase doesn't use foldcase semantics when the haystack is interpolated
samcv have rt's open to track these now
lizmat m: role Foo[::T] { has T $.a }; class Bar {}; use nqp; my $b := Foo[Bar].new; dd nqp::getattr($b,$b.WHAT,'$!a').^name # shouldn't that need to say Bar ? 20:45
camelia "T"
Geth roast: 8773139c37 | smls++ | S06-signature/slurpy-and-interpolation.t
Test interpolating a typed hash into an argument list

Closes RT #126951		 21:05
synopsebot RT#126951 [open]: rt.perl.org/Ticket/Display.html?id=126951 [TESTNEEDED] Interpolating a typed hash into an argument list produces wrong keys
Geth rakudo/nom: 92e51c3d4e | (Elizabeth Mattijsen)++ | 5 files
Compile time defaults for attributes stored as value

  - store the compile-time value of a default if possible
  - makes object creation with concrete compile-time default values faster
   - about 40% faster for a single attribute, like "has $.foo = 42"
  - rather than generating a method to be installed and called during init
  - only supports concrete default values for now, type objects use old behaviour
... (12 more lines)		 21:32
lizmat seems to knock off about .1 second for my run of test-t # [Tux] 21:34
Zoffix \o/ 21:35
Geth roast: 59398cd9cf | smls++ | S06-signature/slurpy-and-interpolation.t
bump plan for commit 8773139c3
 21:36
roast: 9ea27ad2e7 | smls++ | S03-metaops/reduce.t
Add test for reducing a Seq with operator =>

This closes RT #130906.
synopsebot RT#130906 [open]: rt.perl.org/Ticket/Display.html?id=130906 [TESTNEEDED] [REGRESSION] Reducing a Seq with `=>` fails with "P6opaque: no such attribute '$!reified' in type List"
Geth roast: be53acf443 | (Elizabeth Mattijsen)++ | S12-introspection/attributes.t
Adapt tests to new possible Attribute.build values
 21:37
roast: a745629ef4 | (Elizabeth Mattijsen)++ | 2 files
Untodo now passing tests
 21:39
lizmat calls it a day 21:43
yoleaux Zoffix: Blob.gist, bruh 21:45
Zoffix yoleaux: already did it, bruh 21:46
yoleaux Zoffix: geth synopsebot 22:02
Zoffix yoleaux: meh, don't feel like it
gfldex m: my %h = :1a, :2b; %h<a c>:exists:delete; dd %h 22:07
camelia Hash %h = {:b(2)}
gfldex m: my %h = :1a, :2b; say %h<a c>:exists:delete; dd %h
camelia (True False)
Hash %h = {:b(2)}
buggable New CPAN upload: SDL2-Raw-0.1.tar.gz by TIMOTIMO cpan.metacpan.org/authors/id/T/TI/...0.1.tar.gz 23:22
Zoffix AlexDaniel: fail.rakudo.party/t/MOLD 23:40
AlexDaniel_ awesome
travis-ci NQP build errored. Samantha McVey 'Add a removal message if ornate parentheses are used 23:41
travis-ci.org/perl6/nqp/builds/284315409 github.com/perl6/nqp/compare/02a42...6d78eef877
Zoffix AlexDaniel_: also, manager interface now has a trashcan button on tickets. That's for deleting spam tickets
AlexDaniel_ Zoffix: oh. OK! Thanks
Zoffix (well, I guess proper solution is to make it handle them automatically... but I kinda thought of adding a button lol) 23:43
AlexDaniel_ Zoffix: uh, what's the time limit exactly? 1 year? 23:44
AlexDaniel_ didn't understand the magic here github.com/zoffixznet/r6/commit/6d...a40f02R112 23:45
Zoffix If difference of years since "Last updated" is over 1 year, then tag mold
AlexDaniel_ ok!
Zoffix Hm... well, it just considers years themselves... I guess I should change it to just unix time diff 23:46
Zoffix rebuilds again 23:48
timotimo Zoffix: i'm not exactly sure what about the modules list rebuild is so slow?
Zoffix AlexDaniel_: k, now it's marking them if last update is older than 60*60*24*365*2 seconds ago 23:51
timotimo what happens after the "Parsing META file"? it takes like 2 seconds until the next output
AlexDaniel_ buggable: tag MOLD
buggable AlexDaniel_, There are 234 tickets tagged with MOLD; See fail.rakudo.party/t/MOLD for details
AlexDaniel_ that's more approachable :)
Zoffix: thanks! ♥
Zoffix is tired of web stuff and goes to hunt for a rakudo ticket to squash
AlexDaniel_ has spent more than an hour on an LTA ticket :'( 23:52
Zoffix Actually no, I'm gonna eat sleeping pills and play GTA 5 23:53
Gonna squashathon tomorrow :)
Zoffix &
timotimo have fun! :)
Zoffix timotimo: I think it does github API fetching business for new commits after parsing meta + there's METAChecker postprocessor. I'd start looking in those two places 23:54
Zoffix &
timotimo time to nytprof maybe ;) 23:59

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!